41% of Montrealers unsure who to elect mayor in municipal election: poll

While Soraya Martinez Ferrada is the frontrunner, by a slim margin, in the race to become Montreal’s next mayor, most Montrealers simply don’t know who to vote for yet.

That’s according to a new Léger survey commissioned by the Montreal Police Brotherhood.

The poll shows Ferrada, the leader of Ensemble Montréal, tops voter intentions with 18 per cent, ahead of Projet Montréal Leader Luc Rabouin at 15 per cent. In third is Craig Sauvé, the head of Transition Montréal, at seven per cent.

Ten per cent selected “another candidate.” Other candidates are Jean-François Kacou, representing new party Futur Montréal, and Gilbert Thibodeau of Action Montréal.

But the largest proportion of voters are those who are undecided, with 41 per cent of respondents saying they did not know or refused to answer the question.

“With 41 per cent of people undecided, this poll shows that people aren’t enthusiastic about the offer from the two old parties, and above all, that it’s a wide-open race where anyone can win,” Sauvé said in a news release.

Futur Montréal leader Kacou echoed Sauvé, a day after the survey was released, and said, “The old parties have had their chance. Montrealers are tired of broken promises and broken systems. Futur Montréal is the only real alternative that puts citizens first.”

An overwhelming number of respondents also indicated they wanted new faces at city hall, with 61 per cent saying they preferred a new team at the helm. Current Mayor Valérie Plante is stepping down once her second term ends.

The housing and homelessness crisis were the top two issues of concern for survey respondents, followed by road construction, crime and taxes. When choosing between Ferrada, Rabouin and Sauvé, most respondents felt Ferrada was best suited to deal with those issues.

Montrealers head to the polls Nov. 2.

The survey was conducted online with 1,008 respondents between Aug. 13-19, 2025. The results were weighted according to age, gender, district of residence, mother tongue, level of education, presence of children in the household, and residential status to ensure a representative sample of the study population. As a guide, a probability sample of 1,008 respondents would have an assumed margin of error of ±3.09 per cent.